Spaces:
Running
Running
| from dataclasses import dataclass | |
| from enum import Enum | |
| class Task: | |
| benchmark: str | |
| metric: str | |
| col_name: str | |
| # Select your tasks here | |
| # --------------------------------------------------- | |
| class Tasks(Enum): | |
| # task_key in the json file, metric_key in the json file, name to display in the leaderboard | |
| Persian_MMLU = Task("Persian MMLU", "Exact Match", "Persian MMLU") | |
| Persian_Math = Task("Persian Math", "Math Equivalence", "Persian Math") | |
| ParsiNLUـEntailment = Task("ParsiNLU Entailment", "Exact Match", "ParsiNLU Entailment") | |
| FarsTail_Entailment = Task("FarsTail Entailment", "Exact Match", "FarsTail Entailment") | |
| ParsiNLU_Machine_Translation_Fa_En = Task("ParsiNLU Machine Translation Fa En", "English Sentence Bleu", "ParsiNLU Machine Translation Fa-En") | |
| ParsiNLU_Machine_Translation_En_Fa = Task("ParsiNLU Machine Translation En Fa", "Persian Sentence Bleu", "ParsiNLU Machine Translation En-Fa") | |
| ParsiNLU_Reading_Comprehension = Task("ParsiNLU Reading Comprehension", "Common Tokens", "ParsiNLU Reading Comprehension") | |
| Persian_News_Summary = Task("Persian News Summary", "Persian Rouge", "Persian News Summary") | |
| # Your leaderboard name | |
| TITLE = """<h1 align="center" id="space-title">ParsBench Leaderboard</h1>""" | |
| # What does your leaderboard evaluate? | |
| INTRODUCTION_TEXT = """ | |
| <p>This leaderboard is created using <a href="https://github.com/shahriarshm/parsbench">ParsBench</a> framework benchmarking toolkit.</p> | |
| <p>In this leaderboard we evaluate the open-weight LLMs based on the Persian (Farsi) language tasks.</p> | |
| <p><strong>Note:</strong> We've only added the GPT and Claude 3.5 Sonnet models evaluation for a better comparison between the open LLMs and SoTA models.</p> | |
| """ | |
| # Which evaluations are you running? how can people reproduce what you have? | |
| LLM_BENCHMARKS_TEXT = f""" | |
| We evaluated the models on these datasets: | |
| - [Persian MMLU (Khayyam Challenge)](https://huggingface.co./datasets/raia-center/khayyam-challenge) | |
| - [Persian Math](https://github.com/Ipouyall/Benchmarking_ChatGPT_for_Persian) | |
| - [ParsiNLU Entailment](https://huggingface.co./datasets/persiannlp/parsinlu_entailment) | |
| - [FarsTail Entailment](https://github.com/dml-qom/FarsTail) | |
| - [ParsiNLU Machine Translation Fa-En](https://huggingface.co./datasets/persiannlp/parsinlu_translation_fa_en) | |
| - [ParsiNLU Machine Translation En-Fa](https://huggingface.co./datasets/persiannlp/parsinlu_translation_en_fa) | |
| - [ParsiNLU Reading Comprehension](https://huggingface.co./datasets/persiannlp/parsinlu_reading_comprehension) | |
| - [Persian News Summary](https://huggingface.co./datasets/HooshvareLab/pn_summary) | |
| """ | |
| EVALUATION_QUEUE_TEXT = """ | |
| For now, you can contact me at [email protected] for submitting a new request. | |
| """ | |
| CITATION_BUTTON_LABEL = "Copy the following snippet to cite these results" | |
| CITATION_BUTTON_TEXT = r""" | |
| @misc{ParsBench, | |
| title = {ParsBench Leaderboard}, | |
| author = {Shahriar Shariati}, | |
| howpublished = {\url{https://huggingface.co./spaces/ParsBench/leaderboard}}, | |
| year = {2024} | |
| } | |
| """ | |